A Highway Capacity Planning Application and Development of Default Values in North Carolina

Size: px
Start display at page:

Download "A Highway Capacity Planning Application and Development of Default Values in North Carolina"

Transcription

1 A Highway Capacity Planning Application and Development of Default Values in North Carolina Daniel J. Findley, PhD, PE* Senior Research Associate, Institute for Transportation Research and Education North Carolina State University Centennial Campus Box 0, Raleigh, NC () -, () - (fax) Jeffrey C. Chang, EI Research Assistant, Institute for Transportation Research and Education North Carolina State University Centennial Campus Box 0, Raleigh, NC () -, () - (fax) Christopher L. Vaughan, PE Research Assistant, Institute for Transportation Research and Education North Carolina State University Centennial Campus Box 0, Raleigh, NC () -0, () - (fax) Bastian J. Schroeder, PhD, PE Assistant Director of Highway Systems, Institute for Transportation Research & Education (ITRE) North Carolina State University Centennial Campus Box 0, Raleigh, NC () -, () - (fax) Robert S. Foyle, PE Retired Associate Director, Institute for Transportation Research and Education (ITRE) North Carolina State University David M. Alford, PE Manager, Engineering Applications Development North Carolina Department of Transportation () 0-, ()0-00 (fax) November, 0 Submitted for presentation and consideration for publication at the rd Annual Meeting of the Transportation Research Board, January -, 0 Word Count:, text words plus,000 for figures/tables ( * 0) =, total * Corresponding Author TRB 0 Annual Meeting

2 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 ABSTRACT The North Carolina Level of Service (NCLOS) methodology provides a planning-level assessment of transportation facilities through a generally faithful implementation of the 00 Highway Capacity Manual (HCM) to derive service volume estimates and performance expectations, with a few minor deviations from the HCM 00. Implemented in the NCLOS software tool, the method gives a visual representation of traffic volume plotted against the various measures of effectiveness for each facility type. The graphical output shows the feasible performance range from the best to the worst case scenarios for each facility type in North Carolina based on different default values, as well as an average default scenario. The user-defined subject facility is evaluated within that range of values to give the user a direct assessment of the performance. The user can conduct a sensitivity analysis of various scenarios by altering the input values to represent possible design considerations for a particular highway. The tool also enables the user to produce a numerical report detailing the results of the analysis, as well as the ability to export the calculated capacity to transportation planning software and travel demand models. The main contributions of this paper are the NCLOS methodology itself, as well as extensive work on default value development. The paper further provides an example application to an urban street segment case and a two-lane highway facility, with the latter offering interesting insights in the behavior of the three different performance measures in that method. Overall, NCLOS is a powerful tool to allow quick planning-level assessment of the capacities of various road segments. TRB 0 Annual Meeting

3 Findley, Chang, Vaughan, Schroeder, Foyle, Alford INTRODUCTION The Transportation Planning Branch of the North Carolina Department of Transportation (NCDOT) is responsible for working with outside planning agencies in providing engineering and planning assistance for the current, proposed, and potential highway network in North Carolina. This branch is charged with identifying future highway needs through the transportation planning process. This process requires the use of modeling and forecasting techniques to determine potential needs and improvements in the transportation system. Accurate travel demand modeling requires appropriate values for roadway capacities and service volumes at various levels of service (LOS). Tools such as the Highway Capacity Manual (HCM) are valuable for performing detailed analyses of facilities and corridors given a series of input data. However, the scarcity of information typically available at the planning stages, coupled with the relative complexity of the HCM product, make direct use of the HCM impractical or inefficient for forecasting applications. The HCM is primarily designed for operational analyses; traditionally it is not particularly well suited to the reverse process of determining acceptable roadway demands for various maximum service volumes or capacities at LOS thresholds. In early 0, the Highway Capacity Manual (HCM 00) was available for transportation facility analyses (). There are significant and important improvements for many of the methodologies in the HCM 00 based on the most recent national research, including the presentation of planning-level service volume tables. As with previous editions of the manual, HCM 00 is the standard for determining the capacity of most highway facilities in the United States. NCDOT commissioned the development of the North Carolina Level of Service (NCLOS) software, a transportation planning application, based off the Highway Capacity Manual in 00 (). It used the 000 edition s (HCM 000) operational analysis methodologies and service volume calculations to determine capacities for each level of service (LOS), A through E, for freeways, multilane highways, two-lane highways, and arterial streets. When HCM 00 was officially released in spring 0, there were adjustments to the highway facilities methodologies and revisions in analysis; this also included several new facility types, enhanced capabilities, and additional quantitative service measures. As a result, the NCLOS software required updates to programming the HCM methodology, input definitions, and default values. The concepts and considerations taken into account may have use beyond NCDOT s software and could help other practitioners apply the HCM operational analytics for planning applications. BACKGROUND NCLOS is used extensively in planning applications within NCDOT and is delivered through a webbased service to business units at NCDOT. Output capacities are used in travel demand forecasting models and in developing Comprehensive Transportation Plans (CTPs). Output values can also be used in the statewide travel demand model now under development by NCDOT. Currently, the tool is also used to provide data for the Performance Metrics Dashboard and as a scoring component in the Strategic Prioritization Process and Urban Loop Prioritization Process. Practitioners from the NCDOT Prioritization Office implement the tool for analysis required for statewide planning, in addition to engineers and planners inside and outside the agency who desire an automated tool for highway capacity planning. This tool is used primarily as a high-level analysis for making comparison across the state or amongst projects. Detailed operational analysis using the HCM 00 would supersede the planning level results delivered by NCLOS, particularly as more data becomes available that would supplant default values. TRB 0 Annual Meeting

4 Findley, Chang, Vaughan, Schroeder, Foyle, Alford The application is unique in that it provides a graphical display of qualitative level of service (LOS) measures, based from quantitative service measures, plotted against the capacity, or annual average daily traffic (AADT), for a given highway facility under specified conditions. The program uses traffic and roadway parameters, or input values, to produce a capacity table for various conditions. The software has sets of input values for best case, worst case, and default conditions for each highway facility. These capacities are plotted automatically. NCLOS also allows users to define input values, and subsequently plots a curve with the three aforementioned conditions. The user-defined curve shifts relative to the default case depending on the variations to input values, and generally remains between the best case and worst case curves, unless the user has valid reasons to deviate beyond these boundary conditions. The input fields and calculations are derived from the HCM 00 methodologies. NCLOS is applicable for freeway, multilane highway, two-lane highway, and arterial facilities, thus corresponds with Chapter, Basic Freeway Segments; Chapter, Multilane Highways; Chapter, Two-Lane Highways; and Chapter, Urban Street Segments, respectively. Significant changes between the HCM 000 and HCM 00 methodologies are outlined. Basic Freeway Segments Changes to speed-flow curves and the free-flow speed (FFS) equation that affect the analysis of the operational methodology (,,,,) o The procedure recommends using the nearest -mph FFS increment for quantitative analysis, thus eliminating the need for interpolating between any of the pre-defined - mph curves. Multilane Highways Changes to speed-flow curves o The procedure recommends using the nearest -mph FFS increment for quantitative analysis, thus eliminating the need for interpolating between any of the pre-defined - mph curves. Two-Lane Highways Elimination of the bi-directional analysis procedure o HCM 000 provided procedures for one-way analysis for climbing and passing lanes and bi-directional analysis for all other segments; one-way analysis used for mountainous terrain and the latter for level and rolling terrain. The inconsistencies between the two procedures led to the elimination of bi-directional analysis in HCM 00. The one-way capacity of a two-lane highway remains at,00 pc/h consistent with HCM 000, while the two-way capacity is limited at a flow of,00 pc/h. Revisions to basic characteristic curves and tables Introduction of Class III Two-Lane Highway type based on a Florida Department of Transportation analysis procedure (,0,) Urban Street Segments Elimination of the four classes of arterials (I, II, III, and IV) Individual urban street segments can have a signalized, roundabout, or all-way stop-controlled (AWSC) intersection at a segment begin/end point (,,) o Lengthier segments can be divided into unsignalized sections. o The basic signals can be actuate to vary effective green times per cycle unless all cycles are operating at capacity. TRB 0 Annual Meeting

5 Findley, Chang, Vaughan, Schroeder, Foyle, Alford METHODOLOGY NCLOS needed to be updated to incorporate the new methodologies in HCM 00 and other enhancements to remain current with the state-of-practice (). The most critical steps in the methodology implemented in NCLOS was the development of default values, the actual coding and implementation of the HCM 00 methodologies, and the generation of the graphical user output. The development of default values is discussed in this section, while the programming of methodologies and plotting of LOS and AADT is discussed in the Implementation section. Default Values The data used for default value development were based on North Carolina traffic volume data, recommendations from NCHRP Report (), the HCM 00, and professional judgment, which is briefly discussed in the Results section of this paper. This planning level method required the development of many defaults, ranging from hourly, directional, and peak hour factors, to free-flow speed defaults, to more complicated defaults for signal timing and platooned arrivals at intersections. Consequently, a significant effort was required to analyze the North Carolina traffic data. Data were collected by NCDOT beginning in and included information up until 0. Data analysis showed no significant differences between earlier and later data, so all data were included in the full analysis. Detailed data from Automated Traffic Recorder (ATR) stations were obtained from the NCDOT Traffic Surveys Group in the form of a database. The first step of the process was to geocode, traffic volume data points, a step necessary to match volume information with a GIS inventory of roadway facilities and facility types for the state. Of the, points, were automatically geocoded and, required manual geocoding. Another effort required the classification of roadways in terms of their functional classification and area type (urban, suburban, or rural). TABLE details the selection process used to categorize each of the roadway types. At the conclusion of the geocoding, the data points were then classified based on the functional classification shown below using geospatial referencing. Once this was completed, default values specific to North Carolina for each classification were established. This extensive effort ensured that the default values produced in the NCLOS program were accurate based on typical North Carolina traffic volume data. TABLE North Carolina Roadway Classification Definitions Roadway General Type Classification Setting Selection Criteria Selection by Facility Type and Speed Urban Roadway has speed limit of mph or less Limit Arterials Facility is within an Suburban Roadway has speed limit of mph or greater MPO or Smoothed Urbanized Area Rural No roadways exist in this category Boundary Selection by Number No roadways exist in this category (roadway Urban of Lanes classified as Arterial) Two-lane No roadways exist in this category (roadway Highways Suburban Facility is outside an classified as Arterial) MPO or Smoothed Rural Class I = Primary Routes TRB 0 Annual Meeting

6 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 0 Multilane Highways Freeways Urbanized Area Boundary Selection by Facility Type and Area Type Selection by Facility Type and Area Type Urban Suburban Rural Urban Suburban Rural Class II = Secondary Routes Class III = No roadways exist in this category (roadway classified as Arterial) Roadway is within a MPO and a Smoothed Urbanized Area Boundary Roadway is within a MPO, but outside Smoothed Urbanized Area Boundary or vice versa Roadway is outside any MPO and outside any Smoothed Urbanized Area Boundary Roadway is within a MPO and a Smoothed Urbanized Area Boundary Roadway is within a MPO, but outside Smoothed Urbanized Area Boundary or vice versa Roadway is outside any MPO and outside any Smoothed Urbanized Area Boundary RESULTS Following the geocoding and categorization process, the available traffic characteristics of the peak hour factor, hourly K-factor, and directional D-factor were graphed and analyzed. The peak hour factor (PHF) describes the amount of peaking within the peak hour, the K-factor represents the proportion of the total daily traffic that travels during the peak hour, and the D-factor characterizes the directional split of traffic. The figures show a frequency distribution of all data points for a given factor, separated by roadway type, and separated by area type using the definitions for urban, rural, and suburban in TABLE. The corresponding tables in each figure tabulate statistical values of sample size (N), mean, median, and mode for each roadway type. N represents the number of traffic volume data points available for each roadway classification. The data represent the most current count data available at locations across North Carolina, which includes years of observations from sensor stations across the state. Mean is the statistical average, median is the numerical value separating the higher and lower halves of the data, and mode is the value occurring most often. The median values were selected as the default value for each parameter and rounded appropriately, along with national data suggested by the HCM 00 and NCHRP Report. Overall, the values show little variability among the roadway type dimensions considered in this evaluation. The majority of data on PHF presented in FIGURE are within a range from 0. to.0, with an overall mean and median of 0. and 0., respectively. The distinction between urban, suburban and rural facilities generally shows little difference in the distributions. Comparing PHF across facility types, both rural freeways and two-lane highways suggest a more dispersed distribution than arterials and freeways, with a greater frequency of low peak hour factors (traffic more focused within a -minute period). The distribution of the hourly K-factor in FIGURE shows most of the data were contained in a range of 0.0 to 0., with an overall mean and median of 0.0 and 0.0, respectively. Again, the collected data show little difference between facility types and between different area types. Finally, the directional D- factor shows the greatest degree of dispersion with a range of frequently observed values from 0. to 0.. TRB 0 Annual Meeting

7 Findley, Chang, Vaughan, Schroeder, Foyle, Alford The overall mean was 0. and the overall median was 0., as seen in FIGURE. Similar to the other measures, the distributions across facility types and area types were surprisingly similar. As a result of these analyses, the team proposed the use of a common K-factor for all area and facility types as shown in TABLE, even though national defaults from NCHRP Report suggested greater variability. For PHF, the recommendation was a common default of 0.0, except for slightly lower values for rural freeways and two-lane highways. For the D-factor, a common default of 0.0 was proposed, except for a lower value for suburban multilane highways, and a higher value for rural freeways. TRB 0 Annual Meeting

8 Findley, Chang, Vaughan, Schroeder, Foyle, Alford FIGURE PHF factor data. TRB 0 Annual Meeting

9 Findley, Chang, Vaughan, Schroeder, Foyle, Alford FIGURE K factor data. TRB 0 Annual Meeting

10 Findley, Chang, Vaughan, Schroeder, Foyle, Alford FIGURE D factor data. TABLE North Carolina Traffic Characteristic Data Factor Facility Type Urban Suburban Rural NC National* NC National* NC National* Freeway N/A K Multi-lane Highway N/A Two-Lane Highway N/A Arterials 0.0 N/A 0.0 N/A N/A N/A Freeway N/A D Multi-lane Highway N/A Two-Lane Highway N/A Arterials 0.0 N/A 0.0 N/A N/A N/A Freeway N/A PHF Multi-lane Highway N/A Two-Lane Highway N/A 0. N/A N/A 0. N/A Arterials N/A N/A N/A * National data represents recommended defaults from HCM 00 and NCHRP Report TRB 0 Annual Meeting

11 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 In addition to these default values that were supported by field data, various additional defaults had to be developed. In the absence of North Carolina specific data, many of these were based on discussion with NCDOT and traffic engineers, supported by national defaults from the HCM 00 and NCHRP Report as applicable. TABLE shows a subset of these defaults for urban arterials and two-lane highways, with the full list available in the research report (). In addition to NCLOS default values, the table presents program limits, which restrict the user s input to the listed boundary conditions, and practical limits, which guide the user to common input for best and worst case practical values. TRB 0 Annual Meeting

12 Findley, Chang, Vaughan, Schroeder, Foyle, Alford TABLE NCLOS Default Data for Arterials and Two-Lane Highways Program Limits Practical Limits NCLOS Default Value Facility Type Arterials Two-Lane Highway Traffic Factors Roadway Factors Traffic Factors Input Restrict Input Within Boundary Conditions Minimum Maximum Worst Best Alert Users of Uncommon Input Urban Suburban Rural BFFS (mph) K Midsegment Volume (veh/hr) 0 0,000 0,000 User User Other Delays (sec) PHF Platoon Ratio Saturated Flow Rate (per,00,00 lane),00,00,00,00 Startup Time Lost (sec)... Total Delay Due To Turns (sec) Upstream Volume Capacity Ratio Access Points Per Mile Cycle Length (sec) G/C Ratio Intersection Width (feet) Length (feet) 0 00, ,000 0,000 0,000 Length with Restrictive Median (feet) 0 00, ,000,000,000 Number of Lanes (per direction) Proportion With Curb Speed Limit 0 0 D FFS K PHF Percent RVs Percent Trucks/Buses Access Points Per Mile BFFS Terrain Type N/A N/A N/A N/A Level Lane Width, Lateral Clearance 0 0, Length of Grade (miles) Percent Grade Percent No Passing Zones , A Two Lane Class I I I I I * Sources: NC Traffic Volume Data; NCHRP Report ; Previous NCLOS Default Value; Ideal Case or Professional Judgment; A 0% No Passing Zones for Mountainous Terrain with Limited Sight Distance Roadway Factors Source TRB 0 Annual Meeting

13 Findley, Chang, Vaughan, Schroeder, Foyle, Alford IMPLEMENTATION The software implementation of the NCLOS methodology is separated into different modules for each facility type. These modules include Freeways, Multi-lane Highways, Two-Lane Highways, and Arterials. The software was programmed to be consistent with HCM00, but there are several nuances that were incorporated. The operational analysis methodologies for basic freeway segments and multilane highways recommend using the nearest -mph FFS increment for quantification. As plotting LOS versus AADT is a principal feature of NCLOS, the use of -mph increments resulted in step function outputs. To avoid this outcome, where an additional vehicle could significantly change the analysis results, the software utilizes interpolation to determine FFS. Another significant deviation relates to the HCM00 chapter on urban streets. The procedure for operational analysis uses the new signalized intersection methodology in Chapter (); it requires iterative computer calculations and includes a complicated incremental queue accumulation method used to estimate uniform delay at a signal. This method was considered to be too complex for direct implementation in NCLOS. Further, it would have required very detailed signal timing parameters that are not available at the planning-level stage. Consequently, NCLOS uses the Quick Estimation Method for Urban Street Segments found in Chapter 0, Section of HCM00 (,). This methodology provides a more appropriate level of calculation for planning applications. The LOS metric for an Urban Street is based on a measure known as percent free flow speed, which is calculated by the ratio of average travel speed to the free flow speed on the facility. The method was configured to estimate the average travel speed and determine LOS for Arterials. An example of the output from the arterial facility type in NCLOS is shown in FIGURE. This graph presents the user with a set of curves for best, worst, and default cases as well as the current facility, or user-defined curve. As the capacity (AADT) increases, the slope increases and the curve drops off, suggesting that the highway facility operates well at low and moderate volume levels but rapidly deteriorates as volume increases. An assessment of the current facility shows that it performs close to the North Carolina average default. In the planning context, there may be additional opportunities for improvement by adjusting traffic and roadway parameters in order to reach the Best Case scenario. TRB 0 Annual Meeting

14 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 FIGURE Arterial graph from NCLOS. Two lane roads are classified into one of three categories: Class I, Class II, or Class III in accordance to the HCM 00 criteria. Class I is a facility which has motorists who expect relatively high speeds, Class II has motorists who do not necessarily expect high speeds, and Class III serves moderately developed areas. These typology definitions are subjective and can be challenging to apply in practice, particularly on a statewide level. The calculations for the service level of these roads are dictated by the classification designated by the user. However, the resulting accuracy of the outputs from this process subsequently relies on whether the appropriate classification was chosen. The evaluation of each highway class is informed by various measures of effectiveness including percent of the free flow speed, average travel speed, or percent time spent following another vehicle. It should be noted that when using the HCM methodology Class II two-lane highways will yield a higher capacity than that of Class I highways given the same input values. Although counter-intuitive at first examination, the methodology attempts to address driver perception in that expectations are less demanding for Class II routes. Accordingly, the analyst must use judgment in selecting input values for the discrepancies in classification. The NCLOS application allows this comparison between classification type selections by viewing the results for each class. TRB 0 Annual Meeting

15 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 FIGURE presents an example two-lane facility in each class with their associated maximum AADT volumes for various LOS levels (Figure A, Class I; Figure B, Class II; Figure C, Class III). The NCLOS implementation is unique in that it takes all three LOS service measures in the two-lane procedure and plots the results on a common y-axis scale. To accomplish this, note that the y-axis scale had to be reversed for some service measures, where average travel speed (ATS) and percent free-flow speed (%FFS) range from low to high when moving upward on the y-axis scale, while percent time spent following (PTSF) ranges from high to low. The three graphs have further been scaled to show a consistent representation of these service measures. The results suggest that the subject facility performs slightly above the average default for the service measures of PTSF and %FFS in Figures B and C. For the average travel speed measure, the facility performs well below average, and falls below the allowable worst case for low AADT ranges. Interestingly, at a common AADT of 0,000 vehicles per day, the ATS, PTSF, and %FFS measures suggest a LOS of D, D, and C, respectively. This points to challenges when applying the two-lane highway method, where the charts and the resulting LOS can offer different results depending on the class selected. TRB 0 Annual Meeting

16 Findley, Chang, Vaughan, Schroeder, Foyle, Alford A B C FIGURE Two-lane NCLOS output for each class. TRB 0 Annual Meeting

17 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 A comparison of the output for each of the three cases does not indicate which case is most appropriate for the specific roadway facility, since each case is based on assumptions pertaining to that situation. However, it is possible to obtain more information about the relationship of unrelated measures of effectiveness by examining their correlation at LOS thresholds and then examining when each measure controls the LOS output. The lower LOS value should be identified based on either average travel speed or percent time spent following, which can be completed by performing two computations and then selecting the lower LOS value. However, when graphically displaying continuous LOS and AADT curves for both measures of effectiveness, a relationship between the measures must be established. Figure displays the relationship of average travel speed and percent time spent following for a two-lane Class I facility, which was created based on values acquired at each LOS threshold. These measures are inversely related and are not linearly correlated. The values have no physical correlation to each other (i.e. mph travel speed 0% time spent following). This graph does not imply that a driver traveling at mph on a Class I facility would expect to follow another vehicle 0% of the time. Instead, this relationship represents the equivalence of a driver s perception on a Class I facility of traveling at mph or spending 0% of their time following another vehicle. 0 0 Average Travel Speed (mph) Percent Time Spent Following (%) FIGURE Two-lane Class I ATS and PTSF perception equivalence graph. Employing the correlation between the measures of effectiveness, Figure shows a composite two-lane highway analysis graph with both measures and a red line indicating which of the dashed measures of effectiveness lines are controlling the overall LOS. In this example, the percent time spent following measure controls the LOS from,00 to,000 vehicles per day. TRB 0 Annual Meeting

18 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 FIGURE Two-lane highway Class I LOS graph. SUMMARY This section summarizes the maximum capacity outputs calculated from the default values in NCLOS. The output represents the maximum annual average daily traffic volumes, the threshold between LOS E and LOS F, calculated by NCLOS rounded to the nearest hundred. The summary is organized by highway facility, area type, and total number of lanes in both directions. TALBE tabulated capacity outputs were derived by the LOS E performance measures for facilities within the limits of the facility characteristics in North Carolina. TRB 0 Annual Meeting

19 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0 0 TALBE NCLOS Capacity Outputs for LOS E Highway Facility Freeways Multi-lane Highways Two-lane Highways Arterials Maximum Capacity for LOS E under Default Conditions by Highway Facility, Area Type, and Number of Lanes Area Type Total Number of Lanes 0 Urban,0,0, 0,, Suburban,0,,, 0, Rural,,0 0,,0, Urban N/A,,,, Suburban N/A,,,, Rural N/A,0,,0, Urban N/A N/A N/A N/A N/A Suburban N/A N/A N/A N/A N/A Rural, N/A N/A N/A N/A Urban,,,,, Suburban,,,,, Rural N/A N/A N/A N/A N/A * Note: N/A - the combination of highway facility and area type are not applicable by definition or for North Carolina highways; NCDOT default values for urban and suburban area types do not differ for freeways and superstreets CONCLUSIONS AND RECOMMENDATIONS The North Carolina Level of Service (NCLOS) software is a tool developed to implement HCM 00 for planning-level assessment of freeway, multilane highway, two-lane highway, and arterial facilities. The program can be utilized to determine capacities and service volume thresholds. The visual aspect of the program output, aided by the display of best and worst case conditions, allows for an intuitive assessment of the facility performance, and a better interpretation of the subject segment performance relative to the range of expected conditions in North Carolina. The user can attempt various scenarios by altering the input values to represent possible design considerations for a particular highway. Additionally, NCLOS enables users to produce a numerical report detailing the results of the analysis as well as the ability to export the calculated capacity to transportation planning software. By updating input definitions and default values, the procedures have been calibrated to reflect specific observed conditions within the state of North Carolina. However, it is emphasized that default values should always be scrutinized for any new facility, and local adjustments should be made for facilities that fall outside the range of defaults. The analysis of default values generally showed few distinct trends across facility types and geographic region, although various outlier locations were observed in each of the data sets. While the defaults appear to provide a good general representation of expected conditions in North Carolina, user judgment should always be applied to unusual sites. ACKNOWLEDGEMENTS This paper was derived from a study that was funded by the North Carolina Department of Transportation. NCDOT has not evaluated or adopted the findings of this paper or accepted or rejected any of its conclusions. The authors would like to acknowledge the efforts of those who supported the research efforts and enabled the project to be successfully completed. TRB 0 Annual Meeting

20 Findley, Chang, Vaughan, Schroeder, Foyle, Alford REFERENCES. Fain, S.J., C.M. Cunningham, R.S. Foyle, and N.M. Rouphail. NCDOT Level of Service Software Program for Highway Capacity Manual Planning Applications. North Carolina Department of Transportation Report FHWA/NC/00-0. August 00.. Findley, D.J., J.C. Chang, C.L. Vaughan, B.J. Schroeder, and R.S. Foyle. NCLOS Program 00 Update. North Carolina Department of Transportation Report FHWA/NC/0-0. June 0.. Transportation Research Board (TRB), National Research Council. 000 Highway Capacity Manual. Washington D.C., Transportation Research Board (TRB), National Research Council. 00 Highway Capacity Manual. Washington D.C., 00.. Zegeer, J.D., M. Vandehey, M. Blogg, K. Nguyen, and M. Ereti. (00). Default Values for Highway Capacity and Level of Service Analyses. NCHRP Report. Transportation Research Board (TRB), National Research Council.. Schoen, J. A., A. May, W. Reilly, and T. Urbanik. Speed Flow Relationships for Basic Freeway Sections. Final Report, NCHRP Project. JHK & Associates, Tucson, Ariz., May.. Roess, R. Re Calibration of the mi/h Speed Flow Curve and the FFS Prediction Algorithm for HCM 00. Research Memorandum, NCHRP Project. Polytechnic Institute of New York University, Brooklyn, N.Y., January 00.. Hall, F. L. and K. Agyemang Duah. Freeway Capacity Drop and the Definition of Capacity. In Transportation Research Record 0, TRB, National Research Council, Washington, D.C.,, pp... Reilly, W, Harwood, D, Schoen, J, and Holling, M, Capacity and LOS Procedures for Rural and Urban Multilane Highways, NCHRP Project, Final Report, JHK & Associates, Tucson AZ, May Washburn, S.S., McLeod, D.S., and Courage, K.G. Adaptation of the HCM000 Planning Level Analysis of Two Lane and Multilane Highways in Florida, In Transportation Research Record 0, TRB, Washington D.C., 00, pp... Harwood, D. W., I. B. Potts, K. M. Bauer, J. A. Bonneson, and L. Elefteriadou. Two Lane Road Analysis Methodology in the Highway Capacity Manual, Final Report, NCHRP Project 0 (0), Midwest Research Institute, Kansas City Mo., September 00.. Bonneson, J., M. Pratt, and M. Vandehey. Predicting the Performance of Automobile Traffic on Urban Streets: Final Report. NCHRP Project, Transportation Research Board of the National Academies, Washington, D.C., Jan Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB).. Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB).. Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB).. Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB).. Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB).. Highway Capacity Manual (HCM), Chapter. (00). Washington, DC: Transportation Research Board (TRB). TRB 0 Annual Meeting

21 Findley, Chang, Vaughan, Schroeder, Foyle, Alford 0. Highway Capacity Manual (HCM), Chapter 0. (00). Washington, DC: Transportation Research Board (TRB). TRB 0 Annual Meeting